A Fast and Accurate Nonlinear Spectral Method for Image Recognition and Registration

摘要

This article addresses the problem of two- and higher dimensional patternmatching, i.e. the identification of instances of a template within a largersignal space, which is a form of registration. Unlike traditional correlation,we aim at obtaining more selective matchings by considering more strictcomparisons of gray-level intensity. In order to achieve fast matching, anonlinear thresholded version of the fast Fourier transform is applied to agray-level decomposition of the original 2D image. The potential of the methodis substantiated with respect to real data involving the selectiveidentification of neuronal cell bodies in gray-level images.